Meeting Title: [Eden] Daily Standup Date: 2025-07-30 Meeting participants: Fireflies.ai Notetaker Joshua, Robert Tseng, Annie Yu, Cutter Streeby, Amber Lin, Awaish Kumar, Josh, kevinbell, Mitesh Patel
WEBVTT
1 00:01:21.270 ⇒ 00:01:22.360 Annie Yu: Hello, Robert!
2 00:01:23.890 ⇒ 00:01:25.270 Robert Tseng: Hey, Annie?
3 00:01:27.430 ⇒ 00:01:34.300 Robert Tseng: yeah, give me like a minute. I’m responding to a message, and then I know you sent me a couple of screenshots that I haven’t read yet.
4 00:01:35.100 ⇒ 00:01:40.439 Annie Yu: Oh, I think that’s probably for ember.
5 00:01:41.580 ⇒ 00:01:42.719 Annie Yu: Yeah, no worries.
6 00:01:43.220 ⇒ 00:01:43.560 Robert Tseng: Okay.
7 00:02:13.970 ⇒ 00:02:14.760 Cutter Streeby: Yo.
8 00:02:17.120 ⇒ 00:02:17.730 Robert Tseng: Hey!
9 00:02:18.130 ⇒ 00:02:19.220 Cutter Streeby: What up? Bro.
10 00:02:21.040 ⇒ 00:02:28.889 Robert Tseng: Well, I guess the I met with Mattesh and Kevin and I I they kind of disagree with your
11 00:02:29.650 ⇒ 00:02:34.990 Robert Tseng: approach to the Meta purchase event. So yes.
12 00:02:35.390 ⇒ 00:02:38.850 Robert Tseng: yeah. So we’re gonna fix that today.
13 00:02:39.290 ⇒ 00:02:44.219 Cutter Streeby: Yeah. But we need to talk about this shit like sending every purchase event to Meta
14 00:02:44.780 ⇒ 00:02:49.990 Cutter Streeby: is going to show in platform Cpas of $80.
15 00:02:50.440 ⇒ 00:02:51.110 Robert Tseng: Yeah.
16 00:02:51.310 ⇒ 00:02:54.910 Cutter Streeby: Fucking pointless. If we’re judged on N row as
17 00:02:56.860 ⇒ 00:03:00.500 Cutter Streeby: and if they’re saying we can exclude a current client list.
18 00:03:01.380 ⇒ 00:03:05.919 Cutter Streeby: maybe that’ll help. But we already tried that before, and we couldn’t. It.
19 00:03:05.920 ⇒ 00:03:11.160 Robert Tseng: And I don’t think it’s good, then it doesn’t exclude it. Well, the deduplication is not good. So.
20 00:03:11.160 ⇒ 00:03:25.859 Cutter Streeby: And legal, said I couldn’t physically upload the list like theoretically, I could upload a new list every 3 days and update that audience with actual data, but they said that I can’t do that.
21 00:03:26.440 ⇒ 00:03:37.119 Cutter Streeby: So that’s why we had all that random shit built out in segment. That’s like P. 0, and that to fucking purchases of Nad. p. 1. Purchases of Sema.
22 00:03:37.480 ⇒ 00:03:38.100 Robert Tseng: Yeah.
23 00:03:38.610 ⇒ 00:03:45.710 Cutter Streeby: Dude. This is they. This is like somebody is gonna have to
24 00:03:46.090 ⇒ 00:03:50.730 Cutter Streeby: convince me that this is the right way, because this seems fucking. Retarded to me
25 00:03:51.870 ⇒ 00:04:03.100 Cutter Streeby: like what the fuck. The point is to make Facebook a viable channel for new client acquisition, sending a hundred percent of purchases as purchase events to Facebook.
26 00:04:03.430 ⇒ 00:04:11.439 Cutter Streeby: We’re not only going to see 70 glutathione purchases. We’re going to see. 20,000 semaglutide purchases inside of Meta
27 00:04:13.380 ⇒ 00:04:21.300 Cutter Streeby: Cpas are gonna be like 14 bucks and everybody’s gonna be like, spend 10 million dollars. And then everybody’s gonna be surprised when N. Row, as is point 1.
28 00:04:24.480 ⇒ 00:04:25.130 Robert Tseng: Yeah.
29 00:04:26.480 ⇒ 00:04:42.248 Robert Tseng: I mean, I’m sorry. I don’t know how to make the decision there, like I mean what we’ve what we’ve what was currently set in there is based off of your direction. It lines up pretty close to what we have within tableau. It is the new new 1st 1st time purchase orders plus
30 00:04:42.530 ⇒ 00:04:43.650 Cutter Streeby: At the pharmacy.
31 00:04:43.650 ⇒ 00:04:45.400 Robert Tseng: As of the pharmacy. Yeah, so.
32 00:04:45.400 ⇒ 00:04:48.580 Cutter Streeby: If they’re bitching about the sent to pharmacy. Tag.
33 00:04:48.710 ⇒ 00:04:52.740 Cutter Streeby: Okay? Like that is maybe a little too aggressive, but
34 00:04:53.200 ⇒ 00:04:57.930 Cutter Streeby: an actual 1st time purchase. If they cancel or abandon.
35 00:04:58.290 ⇒ 00:05:09.349 Cutter Streeby: it’s gonna do what it does on enroll, as where it’s like we have 800 new patients in tableau. 2 weeks go by, and now we have 750, and the cat goes up
36 00:05:09.920 ⇒ 00:05:14.229 Cutter Streeby: because those cancels and abandons disappear. You know what I mean.
37 00:05:16.090 ⇒ 00:05:24.970 Robert Tseng: Yeah, I mean, I think we probably need another call with Mattesh and Kevin, because they were also insistent that in only doing the 1st time purchase filtering was wrong. So
38 00:05:26.230 ⇒ 00:05:34.059 Robert Tseng: yeah, they they seem. They are insistent that they should be able to exclude previous customers at the from their from their campaigns.
39 00:05:34.736 ⇒ 00:05:37.399 Robert Tseng: I mean, I don’t know like I just said
40 00:05:39.100 ⇒ 00:05:47.529 Robert Tseng: he’s also telling me overrule what Cutter said, like we were about to go make the change today. So we we gotta we gotta be on the same page about this.
41 00:05:48.510 ⇒ 00:05:52.949 Josh: Hold up. Wait a minute. Let me get his. I’m gonna get everybody in here.
42 00:05:52.990 ⇒ 00:05:54.260 Cutter Streeby: Fucking price.
43 00:05:57.900 ⇒ 00:06:00.190 Josh: Kevin Bell. You said too.
44 00:06:01.090 ⇒ 00:06:02.599 Robert Tseng: Yeah, Kevin and Mattesh.
45 00:06:09.180 ⇒ 00:06:09.930 Josh: Okay.
46 00:06:11.220 ⇒ 00:06:18.409 Josh: we will get right to the bottom of this. I don’t know what the fuck you guys are talking about. By the way, I just like just joined so.
47 00:06:18.810 ⇒ 00:06:20.600 Cutter Streeby: Welcome to the party, Buddy.
48 00:06:20.900 ⇒ 00:06:25.039 Cutter Streeby: Yeah, what’s what’s what’s what’s the party? What’s what’s
49 00:06:25.360 ⇒ 00:06:42.850 Cutter Streeby: remember? When we had Fb running with Adm, and they were spending like a million dollars, and it was showing Cpas of like 1, 64. Because we’re sending every single purchase event up there and row, as is point one cash fucking exponentially high.
50 00:06:42.850 ⇒ 00:06:43.450 Josh: Yeah.
51 00:06:44.100 ⇒ 00:06:46.870 Cutter Streeby: That’s what we’re trying to revert to, and.
52 00:06:46.870 ⇒ 00:06:48.629 Josh: No, no, no, no, no.
53 00:06:48.630 ⇒ 00:06:54.019 Cutter Streeby: Yeah, I already said that, but like trying to exclude them from
54 00:06:54.380 ⇒ 00:07:01.939 Cutter Streeby: their thing is like Facebook can exclude from current campaigns. And we tried that, and it doesn’t exclude literally anyone.
55 00:07:03.670 ⇒ 00:07:06.270 Cutter Streeby: So the Cpas in platform will show
56 00:07:06.680 ⇒ 00:07:13.590 Cutter Streeby: 800.
57 00:07:13.980 ⇒ 00:07:16.930 Cutter Streeby: We’re doing great on Facebook. Yeah, you fucking are.
58 00:07:20.180 ⇒ 00:07:21.180 Josh: Okay?
59 00:07:25.900 ⇒ 00:07:26.900 Josh: Oh, no.
60 00:07:27.000 ⇒ 00:07:33.220 Josh: yeah, no, I don’t. Wanna. I wanna make sure it’s 1st purchase event, because we went through and fixed this already once.
61 00:07:36.360 ⇒ 00:07:41.400 Robert Tseng: By the way, Amber, can you send the team to a different room just like you can do your stand up somewhere else? I’ll stay on this phone.
62 00:07:41.400 ⇒ 00:07:43.510 Amber Lin: Oh, okay. Sounds good.
63 00:07:44.060 ⇒ 00:07:45.040 Josh: Good idea.
64 00:07:55.470 ⇒ 00:07:56.320 Amber Lin: Okay.
65 00:08:25.060 ⇒ 00:08:30.130 Robert Tseng: Okay, I’m assuming you’re just gonna drop off and open a different zoom. Like, yeah, yeah, you and Annie should be on this call.
66 00:08:30.130 ⇒ 00:08:31.020 Amber Lin: Okay.
67 00:08:31.180 ⇒ 00:08:31.810 Robert Tseng: Okay.
68 00:08:31.930 ⇒ 00:08:32.610 Robert Tseng: Alright
69 00:08:33.890 ⇒ 00:08:41.530 Robert Tseng: So anyway, like, I’ll just kind of recap before. But if Natasha’s ever come like I think I mean, I could send you the notes. But
70 00:08:44.280 ⇒ 00:09:00.769 Robert Tseng: yeah, they also didn’t like that. We turned off their browser, Pixel and Google Tag manager, because Sebastian persuaded them that we were wrong or something. I still disagree. So we’re they just they just have a lot of confidence in Facebook’s ability to deduplicate, which
71 00:09:01.200 ⇒ 00:09:05.659 Robert Tseng: I don’t know. I think my perspective and Andrew’s perspective is that that’s not true.
72 00:09:06.062 ⇒ 00:09:16.039 Robert Tseng: So anyway, it’s just they just the the short of it. Josh, is that they didn’t like the way that we changed the events we were sending into Meta. So now they want us to kind of
73 00:09:16.330 ⇒ 00:09:27.429 Robert Tseng: revert it back to what it was before. The only thing we’re doing differently is just that we’re sending it from segment instead of Google Tag manager, which I’m not entirely sure. Why, that would be any better. So.
74 00:09:27.810 ⇒ 00:09:43.130 Josh: Well, they’re gonna get to explain themselves, cause they will be in 1 min and 1 min. We’ll get everyone in here, and I will figure out why the fuck we’re on this fucking, stupid track of what the fuck
75 00:09:43.680 ⇒ 00:09:51.880 Josh: join me today on today’s episode of what the fuck like this is. This is my fucking day today, I guess. Just
76 00:09:52.480 ⇒ 00:09:59.369 Josh: come on down, Robert. You can join me on today’s episode of the price is wrong. Bitch.
77 00:09:59.780 ⇒ 00:10:00.300 Cutter Streeby: That’s good.
78 00:10:01.840 ⇒ 00:10:02.760 Josh: Oh, hello!
79 00:10:03.240 ⇒ 00:10:08.050 kevinbell: Learning the re downloading zoom, which is a lovely experience.
80 00:10:09.900 ⇒ 00:10:14.360 Josh: Dude. I bet we’re not a zoom company. This is Robert call Robert.
81 00:10:14.360 ⇒ 00:10:17.600 Josh: Yeah, I’m like, who does zoom? I was playing.
82 00:10:21.280 ⇒ 00:10:24.990 Robert Tseng: And Zoom’s good. It’s got a good Api. We can get a lot of stuff out of it.
83 00:10:25.180 ⇒ 00:10:38.729 Josh: That’s true. That’s true. That’s very so like I I’m I asked everybody to kind of just get in the call, because, like I rob was bringing up a couple of things to me, and it’s I like to hear some of this shit firsthand.
84 00:10:39.690 ⇒ 00:10:47.430 Josh: so I could get a baseline, because, like, we’ve also seen a lot of things in the past and like gone through and seen the movie a few times and.
85 00:10:48.920 ⇒ 00:10:55.229 Robert Tseng: I was just telling them of the changes that we met on yesterday on what you and Mattesh wanted me to change about the track so and.
86 00:10:55.230 ⇒ 00:10:55.610 kevinbell: Yes.
87 00:10:55.610 ⇒ 00:10:59.720 Robert Tseng: Seems to still be disagreement. So I think that’s what we’re wanting to clarify.
88 00:10:59.720 ⇒ 00:11:04.310 Josh: I don’t know if it’s disagree. I just wanna understand what the fuck you guys are trying to accomplish.
89 00:11:04.310 ⇒ 00:11:09.789 Cutter Streeby: So this is Kevin for me, like I got 700 K for new drugs.
90 00:11:10.060 ⇒ 00:11:17.660 Cutter Streeby: If Fb. In platform results are showing me a hundred $78 with 70 purchases.
91 00:11:17.890 ⇒ 00:11:29.110 Cutter Streeby: I’m gonna allocate my money to that channel, but when I check in bask and there’s only 9 fucking purchases, and that is showing 70 with glutathione. That is an actual example.
92 00:11:30.040 ⇒ 00:11:30.770 kevinbell: Yes, we do.
93 00:11:30.770 ⇒ 00:11:35.850 Cutter Streeby: The fucking, worthless platform like. I can’t do anything with it. I can’t spend money. I can’t do anything.
94 00:11:35.850 ⇒ 00:11:51.819 kevinbell: Yeah. So the reason why it wasn’t working it comes down to the fact that when we’re doing view through attribution, and this is all obviously going back to Basque as well. But bask is only doesn’t as we talked about last week or cutter or week before.
95 00:11:51.960 ⇒ 00:11:58.380 kevinbell: bask doesn’t send us Facebook, the data as far as like which product is actually purchased.
96 00:11:58.810 ⇒ 00:12:11.249 kevinbell: that problem is extrapolated even worse when you’re doing view through attribution, so that when someone sees a glut glutathione ad but then purchase a glp. One. Everything is just considered one purchase.
97 00:12:11.380 ⇒ 00:12:16.709 kevinbell: and so there’s some like a double attribution, double purchase that happening right?
98 00:12:17.060 ⇒ 00:12:19.650 kevinbell: So the solution to fix that is a couple of things.
99 00:12:20.090 ⇒ 00:12:25.950 kevinbell: One is just to turn off view through attribution and and have everything be click based. So that way
100 00:12:26.140 ⇒ 00:12:28.220 kevinbell: the click id will determine
101 00:12:28.630 ⇒ 00:12:32.920 kevinbell: which purchase is gonna happen so that will give that will clear up some of the things there.
102 00:12:33.140 ⇒ 00:12:44.789 kevinbell: The other part. That Mattesh and I have talked about in in you. And I have talked about is we’re gonna have multiple pixels on Facebook. So we’ll we’ll
103 00:12:44.920 ⇒ 00:12:49.710 kevinbell: we’ll put a lot of the products in different pixels so that
104 00:12:50.370 ⇒ 00:13:00.950 kevinbell: any purchase that’s gonna happen on some of these like new products. If we do start spending on them on Facebook. They’ll be in this in their own isolated perspective until we get
105 00:13:01.150 ⇒ 00:13:11.740 kevinbell: our own custom, Ehr? And our own custom intake, where at some point we could potentially roll some of these things back into one account, but by separating them into different accounts
106 00:13:12.170 ⇒ 00:13:18.640 kevinbell: and giving them their own pixels. And then just only doing click through attribution.
107 00:13:18.750 ⇒ 00:13:21.090 kevinbell: You’re creating this window
108 00:13:21.460 ⇒ 00:13:29.020 kevinbell: where Facebook can actually operate because you’re right. It was like it was just getting out of control with the amount of view through. And it was just due to the fact that
109 00:13:29.270 ⇒ 00:13:35.309 kevinbell: the the app we’re just not getting Facebook is not getting the data to be able to determine which is which.
110 00:13:35.610 ⇒ 00:13:40.709 Cutter Streeby: So why don’t we send the purchase event? Not from Bask, but from
111 00:13:41.090 ⇒ 00:13:44.260 Cutter Streeby: Robert’s data warehouse, where it actually knows
112 00:13:45.270 ⇒ 00:13:49.530 Cutter Streeby: actual things like this was purchased, sent Facebook.
113 00:13:49.910 ⇒ 00:13:50.390 kevinbell: We were.
114 00:13:50.390 ⇒ 00:13:58.749 kevinbell: The way Robert was describing is that we were doing a ton of other additional filtering, and when you start kind of getting into that
115 00:13:58.920 ⇒ 00:14:01.659 kevinbell: other additional filtering, both
116 00:14:02.183 ⇒ 00:14:09.749 kevinbell: the pixel and the conversion Api setup needs to be aligned. And so when the when you start sending misaligned data.
117 00:14:10.170 ⇒ 00:14:11.430 kevinbell: Then Facebook.
118 00:14:11.610 ⇒ 00:14:20.030 kevinbell: you end up hurting Facebook even more. So from that perspective. So like they, those numbers need to match as much as possible from a deduplication standpoint.
119 00:14:20.030 ⇒ 00:14:29.850 Cutter Streeby: So the reason we opted for an Ft. Purchase event was when we 1st started running on Meta. We were at a million bucks a month, and we were spending on
120 00:14:30.040 ⇒ 00:14:32.980 Cutter Streeby: Sema and Oral Sema. Those are the only 2,
121 00:14:33.590 ⇒ 00:14:38.690 Cutter Streeby: and because we sent every purchase event, including reorders, as purchases
122 00:14:39.210 ⇒ 00:15:00.819 Cutter Streeby: in platform. Cpa. Was like 80 bucks, but true, Cpa. Was like $800. So that’s why we only send Ft. Purchase equals. True, because then, when we were when we look at Nro as and Ncac. Reporting, it’s an actual Ncac. Number. We tried to exclude those groups
123 00:15:00.980 ⇒ 00:15:05.549 Cutter Streeby: like purchasers from Facebook, but it never excluded anybody that I could tell.
124 00:15:05.550 ⇒ 00:15:09.630 Mitesh Patel: So so, cutter, if we have a 7 day click attribution
125 00:15:10.020 ⇒ 00:15:15.260 Mitesh Patel: right? Why would it send and take credit for like renewals? A month later.
126 00:15:15.540 ⇒ 00:15:19.779 Cutter Streeby: Because if you send the purchase event as a purchase event.
127 00:15:20.480 ⇒ 00:15:26.050 Cutter Streeby: those people are gonna have seen those ads or click those ads or done something. And then you’re gonna get
128 00:15:26.190 ⇒ 00:15:30.780 Cutter Streeby: those purchases coming back that are actually just reorders.
129 00:15:31.220 ⇒ 00:15:40.370 Cutter Streeby: That’s why we changed Google, Fb, everything is only trained on ft purchase events so that the algorithm you don’t have to exclude anybody.
130 00:15:40.590 ⇒ 00:15:48.130 Cutter Streeby: You don’t have to exclude current client lists. Anything like that, you still can. But I haven’t seen that work for Meta
131 00:15:48.530 ⇒ 00:16:01.000 Cutter Streeby: when we did this before, and so in Platform Cpas, they oh, I need to renew my order, and they click the button, and they go renew. Same thing was happening on on Vibe.
132 00:16:01.880 ⇒ 00:16:06.229 kevinbell: The thing is so, and that makes sense to an extent. Because
133 00:16:07.580 ⇒ 00:16:17.979 kevinbell: I I get what you’re saying. But at the same time it’s like Facebook should be able to. And, like, you know, all of the products Facebook should be able to determine what is a new customer and what’s not.
134 00:16:18.360 ⇒ 00:16:20.550 kevinbell: I mean, we sent every
135 00:16:21.310 ⇒ 00:16:28.909 kevinbell: you know we didn’t exclude any. I I’ve never had. I’ve never needed to exclude within the integration setup
136 00:16:29.764 ⇒ 00:16:36.820 kevinbell: reordering of of purchases that happen like I’ve never seen Facebook attribute a reorder
137 00:16:37.410 ⇒ 00:16:49.760 kevinbell: to an ad that someone purchased like if if someone purchased an ad 3 months ago, and they’re now in their 3rd month. I’ve never seen that ad all of a sudden. Get 3, you know that 3rd month get 3 3 conversions for that one person
138 00:16:50.340 ⇒ 00:16:55.290 kevinbell: I I can, and maybe that’s maybe that’s something with how bask is set up.
139 00:16:56.135 ⇒ 00:17:05.609 kevinbell: And that might be something that we have to it like, maybe that’s the thing we have to investigate to see if it’s actually doing that. But I’ve never seen it
140 00:17:05.950 ⇒ 00:17:08.350 kevinbell: in a in a place where it’s like all right.
141 00:17:08.720 ⇒ 00:17:13.900 kevinbell: all of these reorders we’re going to attribute to something. I I’ve only seen it attribute
142 00:17:14.460 ⇒ 00:17:18.520 kevinbell: that actual purchase event to a, to an, to an individual product.
143 00:17:18.760 ⇒ 00:17:24.999 Cutter Streeby: So when you, when you started talking, you said there’s an issue with how Bask is sending the purchase event.
144 00:17:25.220 ⇒ 00:17:26.349 Cutter Streeby: How does
145 00:17:27.069 ⇒ 00:17:34.080 Cutter Streeby: how like? How how are we gonna fix that part? Because we have bask still for probably like 3 more months.
146 00:17:34.300 ⇒ 00:17:37.089 Cutter Streeby: Yeah. So in the interim.
147 00:17:37.490 ⇒ 00:17:46.850 Cutter Streeby: how are we gonna make Facebook an actual viable platform for new client acquisition. If we don’t send the actual ft. Purchase event
148 00:17:46.980 ⇒ 00:17:51.160 Cutter Streeby: from our data warehouse to Facebook saying, this is a 1st time purchase.
149 00:17:51.850 ⇒ 00:17:59.369 kevinbell: It’s it’s by separating the customers into their own ad account, or separating the products into their own ad accounts, and then giving them different.
150 00:17:59.370 ⇒ 00:18:01.630 Cutter Streeby: For one drop like even for Sema.
151 00:18:01.630 ⇒ 00:18:02.420 kevinbell: Yeah.
152 00:18:02.420 ⇒ 00:18:11.340 Cutter Streeby: The drug side alone. So we’re running. We’ll run in this ad account. We’ll run Sema med kits and female Hrt, because those are the
153 00:18:11.730 ⇒ 00:18:13.990 Cutter Streeby: groups that make up the weight loss.
154 00:18:14.230 ⇒ 00:18:14.960 kevinbell: Right.
155 00:18:15.460 ⇒ 00:18:23.919 Cutter Streeby: So how do? How do we make Facebook viable again? If we send every single fucking purchase over there.
156 00:18:24.830 ⇒ 00:18:34.719 kevinbell: It’s by. It’s by getting rid of view through attribution. So that way, like any click, it’s only gonna attribute a sale. If someone clicks on an ad and actually goes through.
157 00:18:34.870 ⇒ 00:18:36.540 Mitesh Patel: And within 7 days.
158 00:18:36.540 ⇒ 00:18:37.550 kevinbell: And within 7, th yeah.
159 00:18:37.550 ⇒ 00:18:42.899 Josh: But you guys are missing the hey? If a user who’s already purchased
160 00:18:43.020 ⇒ 00:18:53.270 Josh: actually just clicks one of those ads again just because they see it to log into their portal to place a renewal order. It’s going to give the same amount of credit. That’s the problem.
161 00:18:53.630 ⇒ 00:18:58.520 Josh: That’s the problem that we solved like with the ft or.
162 00:18:59.960 ⇒ 00:19:02.409 kevinbell: It shouldn’t, because it’s like if we’re.
163 00:19:02.410 ⇒ 00:19:09.999 Josh: But it’s doing that like that’s doing that. That’s why you’re in platform numbers look so good. But then, when we’re looking back, it’s so bad.
164 00:19:10.240 ⇒ 00:19:22.269 Josh: It’s bass. It’s a bask issue. I think that fundamentally you guys are right. But it’s like dude. We’re fucked because of the way Basque sucks. And so there is no way to get around it unless we push the ft.
165 00:19:22.530 ⇒ 00:19:23.410 Josh: Metric.
166 00:19:26.080 ⇒ 00:19:30.910 kevinbell: If someone clicks on an ad cause, it should be able to know if someone clicks on an ad if
167 00:19:31.340 ⇒ 00:19:32.159 kevinbell: was already purchased.
168 00:19:32.160 ⇒ 00:19:33.350 Josh: It doesn’t, though it does.
169 00:19:33.350 ⇒ 00:19:33.920 kevinbell: Yeah, but.
170 00:19:33.920 ⇒ 00:19:39.740 Josh: What basket sending baskets are saying. Orders complete unless it’s sending ft. Order complete.
171 00:19:44.130 ⇒ 00:19:47.700 Josh: So like Facebook will treat it all like like, it’s just an order.
172 00:19:49.330 ⇒ 00:19:52.089 kevinbell: So it’s basically combining all orders for.
173 00:19:52.090 ⇒ 00:19:55.100 Josh: Yes, that’s the problem. That is the exact problem.
174 00:19:55.320 ⇒ 00:19:57.459 Josh: which is why it all blows up.
175 00:19:59.590 ⇒ 00:20:04.099 Robert Tseng: Fast sense multiple events per order. If we don’t do our if we don’t use our modeling.
176 00:20:04.300 ⇒ 00:20:07.199 Robert Tseng: So I don’t believe that Meta will be able to deduplicate.
177 00:20:07.200 ⇒ 00:20:24.209 Josh: And this is why Adm got canned. This is why Drip got canned, and how we ended up fixing it for you guys. But now, if you guys want to go back and do the same thing that Adm did and Drip did, we’re going to end up spending 3 million dollars a month on Meta, because it looks awesome, but it blows up all the metrics.
178 00:20:26.310 ⇒ 00:20:27.410 Cutter Streeby: And maybe it’s record.
179 00:20:27.410 ⇒ 00:20:29.809 kevinbell: What were you saying, Robert? Say that again.
180 00:20:30.480 ⇒ 00:20:50.180 Robert Tseng: Yeah. So like the way that the best web hook works, it’ll fire an order. It’ll it’ll fire multiple events per order. It’s it’s a. It’s a range, it’s not. And so the way that Meta deduplication works. Yes, I agree. You need like the Gtm. Event, id or whatever. But if it’s the same order, and it’s multiple events they’re gonna count. It’s I mean, you’re not sending like the same.
181 00:20:50.880 ⇒ 00:20:55.310 Robert Tseng: you’re not set. You’re every. And it’s gonna treat them as separate conversions is basically my point.
182 00:20:55.310 ⇒ 00:21:05.470 kevinbell: So reorders don’t get. So if someone is on, it’s once someone’s on a subscription, it’s 1 order so like new. The next month’s order doesn’t count as a new order. Id.
183 00:21:05.600 ⇒ 00:21:11.309 kevinbell: and because it doesn’t count as a new order. Id. There’s a new event. So like one order, someone’s on.
184 00:21:11.310 ⇒ 00:21:14.860 Cutter Streeby: They bought a quarterly. They don’t get order complete again.
185 00:21:15.100 ⇒ 00:21:18.550 Cutter Streeby: but if they’re on monthly they get order, complete order, complete order, complete.
186 00:21:18.550 ⇒ 00:21:24.320 kevinbell: And it’s it’s just order complete. But it’s the same order. Id. So it’s all that’s going to send to Facebook. So basically, that’s.
187 00:21:24.650 ⇒ 00:21:28.349 Mitesh Patel: Hold hold on. If there, let’s say it’s a monthly conversion
188 00:21:28.690 ⇒ 00:21:33.340 Mitesh Patel: they’re going in through the follow up, not our intake.
189 00:21:34.610 ⇒ 00:21:38.529 Mitesh Patel: Why would we sending any conversion events anywhere in a follow up.
190 00:21:41.090 ⇒ 00:21:43.000 kevinbell: That’s what he that’s what he’s saying.
191 00:21:43.000 ⇒ 00:21:44.190 Mitesh Patel: That’s the problem.
192 00:21:44.190 ⇒ 00:21:45.739 Josh: That is the problem, for sure.
193 00:21:47.600 ⇒ 00:21:50.070 Mitesh Patel: Is it that to?
194 00:21:50.220 ⇒ 00:21:57.440 Mitesh Patel: Isn’t that up to us through the web hooks, or whatever to not send a conversion event in a follow up.
195 00:21:59.580 ⇒ 00:21:59.900 Robert Tseng: I mean.
196 00:21:59.900 ⇒ 00:22:04.369 Mitesh Patel: No Facebook. There’s no Facebook traffic to a follow up
197 00:22:05.050 ⇒ 00:22:10.140 Mitesh Patel: right? A follow up happens when a customer gets an email, they check in whatever.
198 00:22:10.785 ⇒ 00:22:16.830 Mitesh Patel: I’m differentiating. Check in versus follow up, follow up where there’s a check. There’s a transaction right?
199 00:22:17.440 ⇒ 00:22:22.710 Mitesh Patel: That follow up is triggered from an email or or or the or the patient portal.
200 00:22:24.190 ⇒ 00:22:30.049 Mitesh Patel: There’s no way from a Facebook ad or any ad. Does someone go to a follow up?
201 00:22:30.340 ⇒ 00:22:36.830 Mitesh Patel: No, but if they there’s there should be no conversion. Events firing to any system in a follow up.
202 00:22:40.730 ⇒ 00:22:41.669 Josh: Agreed. But that’s.
203 00:22:41.670 ⇒ 00:22:46.429 Cutter Streeby: Wholeheartedly agree with you. That seems like the easy way to fix that shit.
204 00:22:47.030 ⇒ 00:22:47.380 Robert Tseng: Yeah.
205 00:22:47.380 ⇒ 00:22:50.090 Josh: Way to do it. The basket just doesn’t work like that.
206 00:22:50.290 ⇒ 00:22:54.569 kevinbell: So you can’t stop it. You can’t stop it from sending another event.
207 00:22:54.570 ⇒ 00:22:55.500 Mitesh Patel: Fucking, bast.
208 00:22:55.500 ⇒ 00:23:06.999 Robert Tseng: Because a web hook, the way a web hook works a web hook just posts just like it’ll keep sending stuff, whereas an Api you get to pull what you want. We don’t have Api access, so we’re not able to.
209 00:23:07.000 ⇒ 00:23:08.610 kevinbell: This web? Not an Api. Okay.
210 00:23:08.610 ⇒ 00:23:23.889 Robert Tseng: Yeah. So whatever Sebastian set up, and Google tag manager for, the basket is just gonna collect whatever is in the Basque stream, and Baske is sending all this other shit in the stream. And that’s how like. That’s why there’s multiple conversion events.
211 00:23:23.890 ⇒ 00:23:24.230 Mitesh Patel: Alright!
212 00:23:24.230 ⇒ 00:23:27.659 Robert Tseng: Or order, and we and we take care of that at the warehouse. We yeah.
213 00:23:27.660 ⇒ 00:23:34.670 Mitesh Patel: So, Kevin, we don’t. Yeah. Given that. It’s a follow up versus an intake. We don’t want those to go to.
214 00:23:34.850 ⇒ 00:23:35.540 kevinbell: Yeah.
215 00:23:35.850 ⇒ 00:23:38.829 Mitesh Patel: Meta, anyway, for optimization.
216 00:23:39.600 ⇒ 00:23:52.389 Josh: But that’s what’s happening. And that’s again, I’m just as frustrated and just as pissed as you guys. But this is what we fucking originally found out with ATM and then Drip did the same thing. And now I’m watching the same story up for the 3rd time.
217 00:23:52.390 ⇒ 00:23:54.740 Cutter Streeby: And then we have to hack it together.
218 00:23:54.740 ⇒ 00:23:55.230 Josh: Oh!
219 00:23:55.230 ⇒ 00:24:00.200 Cutter Streeby: Around with all this random shit to be able to get usable data out of bass.
220 00:24:02.390 ⇒ 00:24:10.970 Josh: Cameron system is set up the right way. By the way, Cameron has the right setup. So like the way that you guys are talking about doing it will work correctly in Cameron system.
221 00:24:10.970 ⇒ 00:24:11.450 kevinbell: Okay.
222 00:24:11.450 ⇒ 00:24:12.010 Mitesh Patel: Okay.
223 00:24:12.270 ⇒ 00:24:19.439 Mitesh Patel: So filtering for 1st time, I think, makes sense. Given the bask issue.
224 00:24:19.620 ⇒ 00:24:25.290 Mitesh Patel: that’s the only filter we should have. Not the other filters like not prescription going to pharmacy, filter.
225 00:24:26.170 ⇒ 00:24:31.130 Cutter Streeby: Okay, but that’s fine. I don’t disagree. But just so, you know.
226 00:24:31.260 ⇒ 00:24:34.909 Cutter Streeby: it’s gonna fire on order complete. And then when we have
227 00:24:35.850 ⇒ 00:24:42.690 Cutter Streeby: 10 people cancel 15 people refund. That’s also gonna come out of tableau in the backwards.
228 00:24:43.330 ⇒ 00:24:50.870 Mitesh Patel: No, we’re fixing. We’re fixing that cap calculation that calculation will no longer be on confirmed orders, only
229 00:24:51.040 ⇒ 00:24:56.990 Mitesh Patel: it will be on checkout orders. That’s the way marketing campaigns need to be managed.
230 00:24:56.990 ⇒ 00:24:59.559 Cutter Streeby: We look at the report at the end of the month, and.
231 00:24:59.560 ⇒ 00:25:08.769 Mitesh Patel: He, Robert, Robert, and Ann is will be creating a new report for us, so that the cap doesn’t move. Every other day we look at the the report.
232 00:25:08.770 ⇒ 00:25:21.679 Josh: There’s a filter, there’s a filter that’s going to get put in. That’s the correct way to do it, because also you shouldn’t get punished because marketing did their job, you know, marketing like they got the sale like, if, like, we lose them, that’s a retention issue. That’s a different thing.
233 00:25:21.810 ⇒ 00:25:30.248 Mitesh Patel: Yeah, it’s not about marketing. It’s about the campaign or the channel shouldn’t get credit or or or or let you know, decline credit or decrease credit because of it.
234 00:25:30.530 ⇒ 00:25:37.400 Cutter Streeby: We’re gonna send that same order completed event to North Beam as well.
235 00:25:37.660 ⇒ 00:25:39.140 Cutter Streeby: Yeah, alright.
236 00:25:41.480 ⇒ 00:25:48.060 Cutter Streeby: And then maybe we with the new Emr, we can use it like a regular
237 00:25:48.480 ⇒ 00:25:51.599 Cutter Streeby: e-commerce platform, and not have to do this.
238 00:25:52.505 ⇒ 00:25:52.880 Mitesh Patel: Okay.
239 00:25:53.800 ⇒ 00:25:59.430 Josh: That’s the that’s what Robert’s team’s gonna be working on with Cameron from the jump.
240 00:25:59.890 ⇒ 00:26:03.450 Cutter Streeby: Man that’s gonna be a whole different ball game.
241 00:26:03.450 ⇒ 00:26:12.000 Josh: Yeah, dude, I know, I know, I know. Well, anyways, this is this is good. So hopefully, now, everyone is now a hundred percent on the same page of why
242 00:26:12.160 ⇒ 00:26:17.240 Josh: I’m having a Mini freak out, because I’ve literally seen the movie twice, and I don’t want to see.
243 00:26:17.240 ⇒ 00:26:19.689 Mitesh Patel: You don’t want to watch it a 3rd time. Okay.
244 00:26:19.970 ⇒ 00:26:22.169 Josh: 1st time. Filter agreed.
245 00:26:22.800 ⇒ 00:26:24.189 Robert Tseng: Okay, so we’re still gonna change.
246 00:26:24.190 ⇒ 00:26:24.590 kevinbell: Vendors.
247 00:26:24.802 ⇒ 00:26:33.957 Robert Tseng: What we’re saying. If we’re just gonna remove the sent to pharmacy filter, I mean, I’ll send you the list of like I said I would of like what else is in there? I think those are the 2 main things. But
248 00:26:34.170 ⇒ 00:26:34.740 Mitesh Patel: Cool.
249 00:26:35.030 ⇒ 00:26:48.369 Robert Tseng: Yeah. And then, for every order only has one event in our warehouse, like we’ve already kind of cleaned out all the crap there, and it’s always going to use like that. That whole issue that you were talking about, Mattesh, about
250 00:26:49.105 ⇒ 00:26:49.540 Robert Tseng: like
251 00:26:49.680 ⇒ 00:27:03.219 Robert Tseng: transaction values or pricing, pricing, changing, or whatever like we’ve we’ve kind of already dealt with that in the warehouse as well. So we’re always using the 1st one before they apply any random discounts or other stuff that may take away from the value of the order.
252 00:27:03.910 ⇒ 00:27:05.260 Mitesh Patel: Whoa! Hold on!
253 00:27:05.800 ⇒ 00:27:10.720 Mitesh Patel: No, we need to send the order value after discount.
254 00:27:15.110 ⇒ 00:27:16.769 Robert Tseng: For valley, ethnicia.
255 00:27:16.980 ⇒ 00:27:17.780 Mitesh Patel: Yes.
256 00:27:18.290 ⇒ 00:27:26.030 Cutter Streeby: So it’s listed at 2 49, but they get a hundred dollar discount. We want to send the order value to all channels as 1 49,
257 00:27:26.680 ⇒ 00:27:32.180 Cutter Streeby: so that the channel platform shows true. True. Ish Roas.
258 00:27:33.700 ⇒ 00:27:38.239 Robert Tseng: Okay, I will get to clear that up. I need to. I’ll I’ll check that.
259 00:27:38.970 ⇒ 00:27:39.670 Josh: That is very.
260 00:27:39.670 ⇒ 00:27:41.010 Robert Tseng: We need to push that fix. Yeah.
261 00:27:41.010 ⇒ 00:27:46.940 Josh: Very important designation, because right now everything’s overclocking. So all the row as looks over indexed.
262 00:27:47.410 ⇒ 00:27:50.870 Robert Tseng: Yeah, okay.
263 00:27:56.058 ⇒ 00:27:58.361 kevinbell: And then last thing on that Robert
264 00:27:58.880 ⇒ 00:28:06.060 kevinbell: or back to the the conversion Api setup, are we? Gonna is Andrew gonna be able to duplicate this setup, then, for the pixel, as well.
265 00:28:07.090 ⇒ 00:28:18.519 Robert Tseng: Yeah, I think that’s yeah. I mean, it’ll be using the same event size, which, however, I mean, they don’t have the same North being doesn’t have its own happy or whatever. But we’ll be able to
266 00:28:19.070 ⇒ 00:28:21.059 Robert Tseng: fire it into North Cap.
267 00:28:21.060 ⇒ 00:28:34.220 kevinbell: I’m I’m sorry for for Meta, like for for the Facebook pixel, and then the like. We’re you’re making the changes to the the model for the conversion Api setup. But are we going to be able to do something similar for the the pixel.
268 00:28:34.790 ⇒ 00:28:47.429 Robert Tseng: Yeah, I mean on the Gtm side. So I mean, I don’t know if I agreed. I mean, we will do it. But like I’ve Facebook recommends using both of the fail safe option. But like
269 00:28:48.340 ⇒ 00:28:49.420 Robert Tseng: it’s not
270 00:28:49.870 ⇒ 00:29:01.040 Robert Tseng: like, even if we match on event ids, I still think that there we won’t know until we test it. I think it’s possible that you’re still gonna get duplication if we do the same thing on both the pixel and the server.
271 00:29:01.750 ⇒ 00:29:04.160 Mitesh Patel: Alright. But Andrew’s gonna check that out. He’s gonna.
272 00:29:04.160 ⇒ 00:29:09.140 Robert Tseng: He’ll check that. Yeah, yeah, that’s what that’s the risk that we talked about after our call. But yeah.
273 00:29:09.140 ⇒ 00:29:16.609 Mitesh Patel: Yeah, I I think Kevin’s concern, is it? Let’s say we check it. It all looks good that they’re doing the the dedupes properly.
274 00:29:17.050 ⇒ 00:29:21.779 Mitesh Patel: If we’re the pixel right fired by the web hook
275 00:29:22.450 ⇒ 00:29:26.640 Mitesh Patel: was gonna give even repeat orders or renewals
276 00:29:27.640 ⇒ 00:29:35.089 Mitesh Patel: where segment won’t won’t send the the will only send 1st time orders. So there’s gonna be a mismatch.
277 00:29:35.820 ⇒ 00:29:40.639 Mitesh Patel: And if there’s too much mismatch, then Meta is, gonna say, you guys are cheating.
278 00:29:41.070 ⇒ 00:29:46.709 Mitesh Patel: yeah, you know. And and and and the algorithms are gonna be messed up. So I think as much as.
279 00:29:46.710 ⇒ 00:29:47.160 Robert Tseng: Exactly.
280 00:29:47.160 ⇒ 00:29:49.660 Mitesh Patel: That we want that pixel there.
281 00:29:50.316 ⇒ 00:29:55.999 Mitesh Patel: I think because we’re gonna be, we would potentially be sending them different transactions.
282 00:29:56.140 ⇒ 00:29:59.479 Mitesh Patel: We should not send that we should not fire the pixel.
283 00:30:00.770 ⇒ 00:30:08.329 Robert Tseng: Yeah, I mean, we’ll still test both, just to see what the discrepancy is. But I think you’re exactly. I think that’s what we’re I. That’s what I think. We’re gonna see.
284 00:30:08.800 ⇒ 00:30:09.470 Mitesh Patel: Yeah.
285 00:30:10.380 ⇒ 00:30:21.410 Josh: But you can set something up where like, if it’s there’s like a listening, if there’s something listening like if if you know like, that 1st thing fails, then put the backup in like.
286 00:30:21.410 ⇒ 00:30:21.900 Robert Tseng: Yeah.
287 00:30:21.900 ⇒ 00:30:29.100 Josh: Was outside of a range. Then definitely put the backup to failsafe in and tell like, just have some sort of fucking python script running.
288 00:30:29.430 ⇒ 00:30:34.910 Josh: You know what I mean like that shouldn’t be like rockets, and it’s really good. Good companies would do that, anyway.
289 00:30:36.860 ⇒ 00:30:46.179 Robert Tseng: Yeah. Yeah. So I mean, like, like, I say, I mean, if we’re gonna it’ll we would just default to server. And then if there and then we’ll have yeah, like, if there’s
290 00:30:46.850 ⇒ 00:30:51.519 Robert Tseng: I don’t think it ever really worked that way. But, like we would use the pixel as a fail safe.
291 00:30:51.520 ⇒ 00:30:55.380 Josh: You guys can dab stuff. I don’t know. I can’t makes sense.
292 00:30:55.380 ⇒ 00:31:04.170 Robert Tseng: Yeah, yeah, I’m saying, the server is usually the dead safe for the pixel, as I bet we’re kind of using it in reverse this time, which I mean, we’ll we’ll see once we actually Qa, what that? What
293 00:31:04.430 ⇒ 00:31:07.249 Robert Tseng: like, how much, what the discrepancy is.
294 00:31:10.210 ⇒ 00:31:16.099 Josh: Cool any other big thing since I got the whole gang. Get the gang back together. Here.
295 00:31:16.100 ⇒ 00:31:17.030 kevinbell: Sucks.
296 00:31:17.550 ⇒ 00:31:40.249 Josh: Yeah, it does suck dude. Tell me about it. I’m pissed. I want it to work the way that you guys want it to work, too, because that’s how it should be. It’s very simple, but it’s literally just bask is a bunch of fucking children that don’t know how to develop a fucking software and then end up in this scenario where we have to have these stupid workarounds. And if we don’t, I end up spending 2 million dollars to get 12 sales.
297 00:31:43.210 ⇒ 00:31:45.650 Josh: 2.3 2.3
298 00:31:45.650 ⇒ 00:31:48.590 Josh: right 2.3 million dollars. You get 12 sales right?
299 00:31:48.710 ⇒ 00:31:50.559 Josh: And that that can happen.
300 00:31:52.100 ⇒ 00:32:01.830 Cutter Streeby: Oh, just one thing. So, Kevin, we’ve got we we launched a Lp. That combined serm Odt and Serm. Injection into one lp.
301 00:32:01.830 ⇒ 00:32:02.470 kevinbell: Okay.
302 00:32:03.570 ⇒ 00:32:18.140 Cutter Streeby: On Meta, do you? Wanna does it make sense to run them jointly like sermodt ad set serm injection ad set in one campaign and point them at that page, because if that works, then we can roll up and row as
303 00:32:18.310 ⇒ 00:32:21.740 Cutter Streeby: on the dashboard, to be include growth.
304 00:32:21.740 ⇒ 00:32:25.635 kevinbell: Yeah, yeah, we can do that. We can roll them into one
305 00:32:26.320 ⇒ 00:32:36.360 kevinbell: campaign and have them. Or we can. We can do 2 things. We can roll one, we can roll it into one campaign and have them as different ad sets, or we can put them into one ad set and do a catalog ad for the 2 as well. So.
306 00:32:36.400 ⇒ 00:32:38.870 Cutter Streeby: But buck up, Cpas, you think.
307 00:32:40.470 ⇒ 00:32:47.229 kevinbell: If we’re no, because we can do a a cbo on it, and then let.
308 00:32:47.480 ⇒ 00:32:48.190 Cutter Streeby: Cool.
309 00:32:48.340 ⇒ 00:32:51.110 kevinbell: And Facebook figure out which one is is better.
310 00:32:51.460 ⇒ 00:32:52.270 Cutter Streeby: Okay.
311 00:32:52.540 ⇒ 00:33:14.109 Cutter Streeby: sweet. Alright, Robert, that’s what we’re gonna do for enroe as in cerm. So then, now it’s all clear. Now we’re not going to be running independent ads. We don’t have to track them independently, so we can have cerm, and you can hit that little button, and it says, Serm injection. Serm. Odt, you can just split out the ad spends for both equally since then a lot, I said, that’s the easiest way to do it. Yeah, cool.
312 00:33:14.360 ⇒ 00:33:17.160 Cutter Streeby: and then hr same medkit, same.
313 00:33:17.510 ⇒ 00:33:18.060 kevinbell: For.
314 00:33:18.060 ⇒ 00:33:18.400 Robert Tseng: Okay.
315 00:33:18.700 ⇒ 00:33:22.209 kevinbell: For the cerm cutter. Do we want to have the same?
316 00:33:22.670 ⇒ 00:33:32.850 kevinbell: We just wanna have cerm. And then, like, what? What do we do? We want to keep the name. I wanna figure out the naming convention on it on the products, do we? Do we want to get rid of the odt and the injection, and just have cerm, or we still gonna have both.
317 00:33:33.420 ⇒ 00:33:38.660 Cutter Streeby: Can we do that, Robert? Just call it cerm in that naming sheet that we have drifting around.
318 00:33:38.660 ⇒ 00:33:44.420 Robert Tseng: Yeah, yeah, if you could just update it there, like owns that he’ll he’ll finish.
319 00:33:44.420 ⇒ 00:33:44.960 Cutter Streeby: And.
320 00:33:45.410 ⇒ 00:33:48.490 Mitesh Patel: So there’s a $30
321 00:33:48.890 ⇒ 00:33:53.380 Mitesh Patel: aov difference between the Odt and injection.
322 00:33:53.880 ⇒ 00:33:58.869 Mitesh Patel: That doesn’t. I mean, that’s not gonna be too much of an effect on the enroll as right. I mean.
323 00:33:58.990 ⇒ 00:34:10.899 Mitesh Patel: what we’re gonna be, we’re we’re gonna figure out, is a percentage right? Not what percent of people are getting injections which are $30 more, and which percentage are getting the odp, which is.
324 00:34:11.210 ⇒ 00:34:14.419 Mitesh Patel: you know, the 30 and and and and so it’ll be a mixed.
325 00:34:15.929 ⇒ 00:34:23.919 Cutter Streeby: Yeah. And it will also give us a clear indication of like, do people want an injection? Or do they want an oral pill?
326 00:34:24.509 ⇒ 00:34:31.769 Cutter Streeby: Because we’ll still be able to see independent orders inside of tableau. Once you drop that little thing down that Robert builds.
327 00:34:33.400 ⇒ 00:34:34.120 Mitesh Patel: Okay.
328 00:34:40.500 ⇒ 00:34:43.130 Josh: Cool, alrighty Marlon!
329 00:34:47.270 ⇒ 00:34:48.000 Josh: Clear.
330 00:34:48.860 ⇒ 00:34:51.030 Robert Tseng: Alright. Thanks. Guys.
331 00:34:51.420 ⇒ 00:34:52.960 Mitesh Patel: Alright. Thank you. Yeah.